{"slug":"vibeeval-eval-harness","source_name":"vibeeval/eval-harness","name":"Vibeeval/Eval Harness","description":"Formal evaluation framework for Claude Code sessions implementing eval-driven development (EDD) principles","version":1,"lift":{"pass_rate_delta_pts":50,"pass_rate_pct":86.4,"total_cases":22,"passed_cases":19,"tokens_delta_pct":22.7,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-08-03T21:42:05.875567+00:00"},"skill_score":0.8636,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":50,"with_pass_pct":86.4,"without_pass_pct":36.4,"tokens_delta_pct":22.7,"turns_delta_pct":0,"total_cases":22,"cases_aggregated":22,"verdict":"mixed","never_hurt":true,"completed_at":"2026-08-03T21:42:05.875567+00:00","run_id":"50a64d23-22aa-4f4f-b1d2-c667578d3fcd","version_number":1,"is_latest_version":true,"gate":null}],"trust":{"skill_safety":"passed","safety_status":"clean","intent_verdict":"safe","content_status":"clean","indexable":true},"license":"MIT","install_count":0,"manifest_hash":"fcc6062b001c506fcd1ba6c9519c9668df6de9202ad77643d1a092d27983f235","raw_url":"https://app.decimal.ai/s/vibeeval-eval-harness/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/vibeeval-eval-harness"}